PROPOSAL OF A PRIORITY RANKING STUDY METHOD FOR PUBLIC HOUSING REORGANIZATION PLANNING USING UNSUPERVISED MACHINE LEARNING

Abstract

A method was developed to create a relative evaluation of public housing reorganization projects, and the following results were obtained for Osaka Prefectural Housing. Principal component analysis showed that the data for Osaka Prefectural Housing can be reduced to five components. Cluster analysis showed that Osaka prefectural housing can be divided into 10 clusters, and that the characteristics of each cluster can be compared in a relative manner. Using the classification results, we proposed two priorities for project implementation for the Osaka prefectural housing to be reorganized in the next 10 years.